Chapter 2 Discrete-Time Market Model
The single-step model considered in Chapter1 is extended to a discrete-time model with N + 1 time instants t = 0, 1, ... , N. A basic limitation of the one-step model is that it does not allow for trading until the end of the first time period is reached, while the multistep model allows for multiple portfolio re-allocations over time. The Cox-Ross-Rubinstein (CRR) model, or binomial model, is considered as an example whose importance also lies with its computer implementability.
2.1 Discrete-Time Compounding ...... 51 2.2 Arbitrage and Self-Financing Portfolios ...... 54 2.3 Contingent Claims ...... 60 2.4 Martingales and Conditional Expectations ...... 65 2.5 Market Completeness and Risk-Neutral Measures 71 2.6 The Cox-Ross-Rubinstein (CRR) Market Model . . 73 Exercises ...... 78
2.1 Discrete-Time Compounding
Investment plan
We invest an amount m each year in an investment plan that carries a constant interest rate r. At the end of the N-th year, the value of the amount m invested at the beginning of year k = 1, 2, ... , N has turned into (1 + r)N−k+1m and the value of the plan at the end of the N-th year becomes
N X N−k+1 AN : = m (1 + r) (2.1) k=1
51
" This version: July 4, 2021 https://personal.ntu.edu.sg/nprivault/indext.html N. Privault
N X = m (1 + r)k k=1 (1 + r)N − 1 = m(1 + r) , r hence the duration N of the plan satisfies
1 rAN N + 1 = log 1 + r + . log(1 + r) m
Loan repayment
At time t = 0 one borrows an amount A1 := A over a period of N years at the constant interest rate r per year. Proposition 2.1. Constant repayment. Assuming that the loan is completely repaid at the beginning of year N + 1, the amount m refunded every year is given by
r(1 + r)N A r m = = A. (2.2) (1 + r)N − 1 1 − (1 + r)−N
Proof. Denoting by Ak the amount owed by the borrower at the beginning o of year n k = 1, 2, ... , N with A1 = A, the amount m refunded at the end of the first year can be decomposed as
m = rA1 + (m − rA1), into rA1 paid in interest and m − rA1 in principal repayment, i.e. there remains
A2 = A1 − (m − rA1) = (1 + r)A1 − m, to be refunded. Similarly, the amount m refunded at the end of the second year can be decomposed as
m = rA2 + (m − rA2), into rA2 paid in interest and m − rA2 in principal repayment, i.e. there remains
A3 = A2 − (m − rA2) = (1 + r)A2 − m 52 " This version: July 4, 2021 https://personal.ntu.edu.sg/nprivault/indext.html Notes on Stochastic Finance
= (1 + r)((1 + r)A1 − m) − m 2 = (1 + r) A1 − m − (1 + r)m to be refunded. After repeating the argument we find that at the beginning of year k there remains
k−1 k−2 Ak = (1 + r) A1 − m − (1 + r)m − · · · − (1 + r) m k−2 k−1 X i = (1 + r) A1 − m (1 + r) i=0 1 − (1 + r)k−1 = (1 + r)k−1A + m 1 r to be refunded, i.e.
m − (1 + r)k−1(m − rA) A = , k = 1, 2, ... , N. (2.3) k r
We also note that the repayment at the end of year k can be decomposed as
m = rAk + (m − rAk), with k−1 rAk = m + (1 + r) (rA1 − m) in interest repayment, and
k−1 m − rAk = (1 + r) (m − rA1) in principal repayment. At the beginning of year N + 1, the loan should be completely repaid, hence AN+1 = 0, which reads
1 − (1 + r)N (1 + r)N A + m = 0, r and yields (2.2). We also have
A 1 − (1 + r)−N = . (2.4) m r and
1 m log(1 − rA/m) N = log = − . log(1 + r) m − rA log(1 + r)
53 " This version: July 4, 2021 https://personal.ntu.edu.sg/nprivault/indext.html N. Privault
Remark: One needs m > rA in order for N to be finite.
The next proposition is a direct consequence of (2.2) and (2.3). Proposition 2.2. The k-th interest repayment can be written as
N−k+1 1 X −l rAk = m 1 − = mr (1 + r) , (1 + r)N−k+1 l=1 and the k-th principal repayment is
m m − rAk = , k = 1, 2, ... , N. (1 + r)N−k+1
Note that the sum of discounted payments at the rate r is
N X m 1 − (1 + r)−N = m = A. (1 + r)l r l=1 In particular, the first interest repayment satisfies
N X 1 −N rA = rA1 = mr = m 1 − (1 + r) , (1 + r)l l=1 and the first principal repayment is m m − rA = . (1 + r)N
2.2 Arbitrage and Self-Financing Portfolios
Stochastic processes
A stochastic process on a probability space (Ω, F, P) is a family (Xt)t∈T of random variables Xt : Ω −→ R indexed by a set T . Examples include: • the one-step (or two-instant) model: T = {0, 1},
• the discrete-time model with finite horizon: T = {0, 1, ... , N},
• the discrete-time model with infinite horizon: T = N,
• the continuous-time model: T = R+.
54 " This version: July 4, 2021 https://personal.ntu.edu.sg/nprivault/indext.html Notes on Stochastic Finance
For real-world examples of stochastic processes one can mention:
• the time evolution of a risky asset, e.g. Xt represents the price of the asset at time t ∈ T .
• the time evolution of a physical parameter - for example, Xt represents a temperature observed at time t ∈ T . In this chapter, we focus on the finite horizon discrete-time model with T = {0, 1, ... , N}.
0 1 2 3 N
Asset price modeling
The prices at time t = 0 of d + 1 assets numbered 0, 1, ... , d are denoted by the random vector (0) (1) (d) S0 = S0 , S0 , ... , S0 in Rd+1. Similarly, the values at time t = 1, 2, ... , N of assets no 0, 1, ... , d are denoted by the random vector